Market analysis and segmentation

Business analytics

Doing business geographically

• Let’s say you want to open a new franchise for your company. What do you need to know?
  
• Brainstorm!

Doing business geographically

• Activities (among others) might include:
  • Geodemographic segmentation
  • Trade area analysis
  • Analysis of competitors’ locations

Geodemographic segmentation

• The partitioning of geographic units (e.g. households, counties) into groups based on common characteristics, then “naming” those groups
• Examples: Esri Tapestry segmentation; CDRC segmentation (UK)

How segmentation works

• Most common method for geodemographic segmentation: k-means clustering
• Goal: to partition individual units into k groups based on attribute similarity

How k-means clustering works

• k cluster “centers” are chosen for a given distribution of data (either a priori by the analyst or at random)
• Points are assigned to a cluster based on proximity in data space
• Cluster centers are then re-calculated, and points re-assigned based on proximity to the new centers
• The algorithm “converges” when data points stop switching clusters
• Graphical illustration: http://www.naftaliharris.com/blog/visualizing-k-means-clustering/

K-means clustering in ArcGIS

• Available via the “Grouping Analysis” tool in the Spatial Statistics toolbox
• Let’s try it out!

Trade area analysis

• Analysis of characteristics of area from which a business generates a majority of its customers
• Trade areas can be defined by:
  • Euclidean or network distance
  • Customer geography
  • Gravity modeling
  • Common sense

Business Analyst

• Business Analyst for Desktop: comprehensive add-on to ArcGIS with detailed demographic & business data along with modeling functionality
  • Expensive!
• Business Analyst Online: Web-based interface to Business Analyst; less functionality than BA Desktop but very intuitive
• Let’s try it out! https://bao.arcgis.com/esriBAO/login/

Trade area analysis in BAO

Competitor location analysis in BAO